The classical clinical features of haemochromatosis -diabetes, bronze pigmentation of the skin, and liver cirrhosis - have been recognized as a distinct clinical and pathological entity since the work of Trousseau and Troisier in the 19th century. Haemochromatosis is an iron storage disorder (OMIM 235200) that may be secondary to another disease such as thalassaemia, or constitute a primary illness with no underlying disorder. In 1935 Sheldon proposed that primary haemochromatosis was an 'inborn error of metabolism' resulting in excess deposits of iron in tissues (Sheldon 1935). The idea that heredity should play an important role in the disease remained controversial but evidence accumulated that not only was an autosomal recessive mode of inheritance likely but that there was association with genetic polymorphism in the MHC, in particular HLA-A3 (Simon et al. 1976, 1977). There were data to link hereditary haemochromatosis to within 1-2 cM of HLA-A, but finding the actual gene or specific polymorphism proved frustrating until the work of Feder and colleagues in 1996 identified the likely gene and causative mutation (Feder et al. 1996).
Feder and coworkers adopted a positional cloning approach (Section 2.3) to identify the gene underlying hereditary haemochromatosis (Feder et al. 1996). They were able to define a 250 kb candidate region within the 8 Mb of sequence they took as their starting point by using linkage disequilibrium mapping and haplotypic analysis. Some 45 polymorphic markers were analysed in 101 patients with haemochromatosis and 64 controls. The allele having the highest excess frequency in affected individuals versus controls was defined as 'ancestral'. This allowed reconstruction of the ancestral haplotype on which the authors proposed the causative mutation(s) occurred. A genomic position of maximum linkage disequilibrium was mapped that colocalized to a region they identified by analysis of likely historic recombination events on chromosomes bearing the ancestral haplotype.
This 250 kb candidate region was then carefully analysed to identify genes within it by cDNA selection, exon trapping, and genomic DNA sequencing (Feder et al. 1996). This revealed 12 histone genes and three novel genes of unknown function. The products of reverse transcription polymerase chain reaction (RT-PCR) amplified RNA and PCR amplified genomic DNA were then sequenced to try and identify any likely causative mutations in coding DNA of these genes. Two patients with haemochromatosis who were homozygous for the ancestral allele were compared to two control individuals. Eighteen sequence variants were found of which three were nonsynonymous and resulted in an amino acid change in the encoded protein. Two of these lay in the histone H1 gene but one was found in a novel MHC class I-like gene, cDNA 24, now denoted the HFE gene.
Feder and colleagues had found the mutation underlying classical hereditary haemochromatosis. In all four patient chromosomes there was a G to A transition at nucleotide 845 of the open reading frame (ORF) which caused a cysteine to tyrosine substitution at amino acid 282 of the encoded protein (rs1800562, c.845G>A, p.C282Y) (Feder et al. 1996). In their study population, the p.C282Y variant correlated with the presence of the ancestral haplotype. It was present on 85% of haemo-chromatosis chromosomes and 3.2% of controls, giving a carrier frequency of 6.4% in their cohort. The nucleotide substitution resulted in a change in a highly conserved residue involved in intramolecular disulphide bridging in other MHC class I proteins and potentially affected the interaction with P2 microglobulin (Fig. 12.10). Of the 178 patients analysed by Feder and colleagues, 148 were homozygous for p.C282Y while nine were heterozygous. They analysed the non-ancestral chromosome present in these heterozygotes and identified a further C to G variant in exon 2 of the gene which results in a histidine to aspartic acid amino acid substitution at position 63 (rs1799945, c.187C>G, p.H63D) (Feder et al. 1996).
Subsequent biochemical analysis has shown that the HFE protein does not bind iron but through interaction with transferrin receptor 1 (TfR1) it can facilitate cellular uptake of transferrin-bound iron. HFE is an MHC class I-like protein whose ancestral peptide binding groove is too narrow for antigen presentation. Like other class I molecules, however, HFE still requires interaction with P2 microglobulin for its cell surface expression. The p.C282Y variant, by disrupting the disulphide bond critical for binding to P2 microglobulin, impairs HFE stability, transport, and cell surface expression, thus compromising the interaction of HFE with TfR1 (Waheed et al. 1997). Further insights were also gained from studies in mice (Box 12.11).
Primary haemochromatosis is one of the commonest inherited diseases described among people of northern European ancestry. The p.C282Y variant is responsible for the most common form of primary haemochromatosis but the situation is complex, with homozygosity for p.C282Y characterized by low penetrance such that while individuals may be predisposed to the severe pheno-type of iron storage disorder it is very difficult to predict whether, and to what extent, the genetic risk will be manifested (Pietrangelo 2004). Mutations of other iron genes have also been identified and found to be involved in 'non-HFE' haemochromatosis such as TfR2 (encoding transferrin receptor 2), HAMP (hepcidin), HJV (haemoju-velin), and FPN (ferroportin).
The p.C282Y mutation is thought to have arisen relatively recently, perhaps 60-70 generations ago, based on linkage disequilibrium analysis (Ajioka et al. 1997). A worldwide study of 2978 people in 1997 showed that the variant was most frequent in northern European populations and did not occur in Africans, Asians, and native Australians (Merryweather-Clarke et al. 1997). Further studies suggest that the mutation occurred by chance in a single ancestor of Celtic or Viking descent
a heavy chain
Cysteine to tyrosine substitution at position 282 (p.C282Y)
Was this article helpful?